The ability to clean up black white images remains an important task for industries such as banking where the use of black white check imaging is common. By cleaning up such images, unwanted background matter can be eliminated, making the document more readable by both humans and machines. Current techniques include optimization of thresholding when converting the image from grayscale to black white, use of a mask filter that surround regions to be cleaned, and morphological image techniques. However, techniques are still sought that can provide high performance with limited computational overhead. Accordingly, a need exists for a more robust method of cleaning black white images.